R

Ss

La

b

c

a

ARRAA

KARRS

1

mrb(NiKsrNe

h0

Journal of Plant Physiology 212 (2017) 94–104

Contents lists available at ScienceDirect

Journal of Plant Physiology

journa l h om epage: www.elsev ier .com/ locate / jp lph

esearch paper

patio-temporal dynamics in global rice gene expression (Oryzaativa L.) in response to high ammonium stress

i Sun a, Dongwei Di a, Guangjie Li a, Herbert J. Kronzucker b,c, Weiming Shi a,∗

State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, ChinaDepartment of Biological Sciences, University of Toronto, 1265 Military Trail, Toronto, ON, M1C 1A4, CanadaSchool of BioSciences, The University of Melbourne, Parkville, VIC 3010, Australia

r t i c l e i n f o

rticle history:eceived 7 January 2017eceived in revised form 20 February 2017ccepted 21 February 2017vailable online 22 February 2017

eywords:mmonium (NH4

+) toxicityice (Oryza sativa L.)NA-Seq (Quantification) profilingpatial-temporal specificity

a b s t r a c t

Ammonium (NH4+) is the predominant nitrogen (N) source in many natural and agricultural ecosystems,

including flooded rice fields. While rice is known as an NH4+-tolerant species, it nevertheless suffers NH4

+

toxicity at elevated soil concentrations. NH4+ excess rapidly leads to the disturbance of various physiolog-

ical processes that ultimately inhibit shoot and root growth. However, the global transcriptomic responseto NH4

+ stress in rice has not been examined. In this study, we mapped the spatio-temporal specificityof gene expression profiles in rice under excess NH4

+ and the changes in gene expression in root andshoot at various time points by RNA-Seq (Quantification) using Illumina HiSeqTM 2000. By comparativeanalysis, 307 and 675 genes were found to be up-regulated after 4 h and 12 h of NH4

+ exposure in the root,respectively. In the shoot, 167 genes were up-regulated at 4 h, compared with 320 at 12 h. According toKEGG analysis, up-regulated DEGs mainly participate in phenylpropanoid (such as flavonoid) and aminoacid (such as proline, cysteine, and methionine) metabolism, which is believed to improve NH4

+ stresstolerance through adjustment of energy metabolism in the shoot, while defense and signaling pathways,

guiding whole-plant acclimation, play the leading role in the root. We furthermore critically assessedthe roles of key phytohormones, and found abscisic acid (ABA) and ethylene (ET) to be the major regula-tory molecules responding to excess NH4

+ and activating the MAPK (mitogen-activated protein kinase)signal-transduction pathway. Moreover, we found up-regulated hormone-associated genes are involvedin regulating flavonoid biosynthesis and are regulated by tissue flavonoid accumulation.

© 2017 Elsevier GmbH. All rights reserved.

. Introduction

In plants, nitrate (NO3−) and ammonium (NH4

+) are the twoajor inorganic nitrogen (N) forms that are directly absorbed by the

oot system. Even though NH4+ is frequently more readily absorbed

y roots and its assimilation requires less energy than that of NO3−

Kronzucker et al., 2000), only a few species perform well whenH4

+ is the sole N source, while most of them suffer NH4+ toxic-

ty when exposed to excess NH4+ as the sole N source (Britto and

ronzucker, 2002). NH4+ toxicity is typically visible in stunted root

ystems and in chlorosis of leaves, accompanied by dramaticallyeduced tissue accumulation of mineral cations and increased root

H4

+ fluxes (Britto et al., 2001; Britto and Kronzucker, 2002; Lit al., 2014).

∗ Corresponding author.E-mail address: wmshi@issas.ac.cn (W. Shi).

ttp://dx.doi.org/10.1016/j.jplph.2017.02.006176-1617/© 2017 Elsevier GmbH. All rights reserved.

Knowledge of NH4+ toxicity has greatly expanded in recent

years, in particular in terms of signaling transmission and molecu-lar and physiological targets in Arabidopsis (Arabidopsis thaliana)(Li et al., 2014; Bittsánszky et al., 2015). Many NH4

+ toxicity- asso-ciated genes have been identified, such as HSN1/VTC1, GAS1/ARG1,XBAT32, ETO1, DPMS1, AMOS1/EGY1, AMOS2, CAP1, SLAH3, andGln1.2, all of which were associated with sensitivity to NH4

+, whileTRH1, AMT1.3, AUX1, and ETR1 were shown as critical to resistanceto NH4

+ (Bai et al., 2014; Li et al., 2014; Zheng et al., 2015; Guanet al., 2016). Some analyses of NH4

+ toxicity at the whole genome-wide level have also been made. For instance, Yang et al. (2015a)revealed a coordinated regulation of carbohydrate and aminoacid metabolism under high NH4

+ for short periods (up to severalhours) of exposure in rice shoots and roots. Furthermore, Yanget al. (2015b) showed that NH4

+-induced nutritional and metabolicimbalances can be partially overcome by elevated root auxin levels

in Arabidopsis under high NH4

+. Wang et al. (2016) carried outtranscriptomic and physiological analyses of common duckweedin response to NH4

+ toxicity to elucidate the participation of

t Phy

ommorwaesispairotai

eoKstewbNfi1ecsah2arolhosa(ipi

dpcectcf1ei((br

L. Sun et al. / Journal of Plan

xidative damage, lignin biosynthesis, phenylpropanoidetabolism, and programmed cell death (PCD). However, theajority of NH4

+ toxicity research has focused on one gene orne signaling pathway at a time and typically only examined theoot or shoot in isolation. Much less attention has been paid tohole-plant analysis, signaling transduction from root to shoot,

nd feedback signaling from shoot to root. Another infrequentlyxamined aspect is the time course of toxicity development. Due touch restrictions, critical adjustments in signaling and metabolismn the context of whole-plant acclimation and adaptation to NH4

+

tress may have remained invisible. It is well known that stress-riming can enable plant cells to sense very low levels of a stimulusnd allow for initiation of resistance systems when challengednitially, thereafter enabling long-term adaptations to chronic orecurring stress (Conrath, 2011; Conrath et al., 2015). Mappingut the early stress response is, therefore, also critical. Thus, withhe goal of characterizing regulatory networks, a spatio-temporalnalysis of the key changes incurred as a result of NH4

+ exposures needed.

Rice is an NH4+-tolerant model crop known to grow well in soil

nvironments characterized by high NH4+ and has indeed been

ften been described as an NH4+ specialist (Kronzucker et al., 1999;

ronzucker et al., 2000). Nevertheless, even rice has been shown touffer toxicity at elevated levels of NH4

+, albeit at a much higherhreshold than other cereals such as wheat or barley (Szczerbat al., 2008; Balkos et al., 2010; Britto et al., 2014). Rice also offersell-established genetic resources for the application of systems

iology (Chandran and Jung, 2014). Some genes related to excessH4

+ were previously identified in rice: Xie et al. (2015) identi-ed OsSE5 (HO1), which is involved in the heme-heme oxygenase(heme-HO1) antioxidant system and may improve tolerance toxcess NH4

+ by optimizing antioxidant defense. Moreover, OsPTR6ould increase rice growth by increasing NH4

+ transporter expres-ion and GS activity under high NH4

+ supply (Fan et al., 2014). Inddition, OsAMT1.1 transgenic rice has shown superior growth andigher yield under elevated NH4

+ conditions (Ranathunge et al.,014). Recently, Ma et al. (2016) showed GABA (�-aminobutyriccid) can alleviate NH4

+ toxicity by limiting NH4+ accumulation in

ice. However, NH4+ toxicity in rice remains poorly characterized

verall, and examinations both at the whole-plant and subcellularevel are clearly required (Bittsánszky et al., 2015). On the otherand, mechanisms of NH4

+ toxicity in rice seem to differ fromther plants such as Arabidopsis, which is a strongly NH4

+-sensitivepecies, or species such as the common duckweed, whose frondsnd roots are directly exposed to NH4

+ in their natural environmentWang et al., 2016). Thus, identifying the molecular and physiolog-cal mechanisms in the rice response to excess NH4

+ are importantrerequisites to helping crop biotechnologists identify traits for

mproving rice germplasms that perform well at high NH4+.

Because of the importance of rice as a global commodity, as theominant calorie source directly consumed by humans, and theotential threat to its production deriving from elevated soil NH4

+

ontent in heavily fertilized soils, a better understanding of rice tol-rance to excess NH4

+ is clearly needed. To this end, we undertook aomprehensive analysis of global changes in the rice genome duringhe stages of NH4

+ toxicity development using RNA-Seq (Quantifi-ation). We produced RNA-Seq data from shoots and roots sampledrom two-week-old seedlings that were initially grown for 4 h and2 h and supplied with 7.5 mM (NH4)2SO4, which allowed us toxamine the response to the onset of excess NH4

+ that may be invis-ble to analyses conducted at only one time point. Gene OntologyGO) enrichment and Kyoto Encyclopedia of Genes and Genomes

KEGG) pathway enrichment were used to identify the importantiological processes and crucial signaling pathways operating inice shoot and root tissues.

siology 212 (2017) 94–104 95

2. Materials and methods

2.1. Plant growth conditions and stress treatments

Seeds of Oryza sativa ssp. japonica were surface-sterilized with1% sodium hypochlorite for 10 min, washed extensively with dis-tilled water, and then germinated in distilled water at 28 ◦C for2 days. Germinated seeds were transferred into a growth chamber(16/8 h (28/25 ◦C) day/night, a light intensity of 400 �mol m−2 s−1,and a constant relative humidity of 70%) in 4.0-l vessels containingaerated modified Johnson’s solution (2 mM MgSO4; 1 mM CaCl2;0.5 mM KCl; 0.3 mM NaH2PO4; 0.1 mM Fe-EDTA; 20 �M H3BO3;9 �M MnCl2; 1.5 �M CuSO4; 1.5 �M ZnSO4; 0.5 �M Na2MoO4,pH 5.5), either under N-sufficient [1 mM (NH4)2SO4] or N-excess[2.5 mM, 5 mM, 7.5 mM, 10 mM and 20 mM (NH4)2SO4)] condi-tions for an additional 14 days. Then seedling length was measured,and fresh biomass of shoot and root was determined. Solutionswere exchanged every 12 h to ensure that plants remained at anutritional steady state in the hydroponic system. For the high-NH4

+ (HN) treatment, seedlings were first grown on N-sufficientsolution for 14 days. Then, seedlings were transferred to an N-starvation solution (no N) for another 3 days to deplete N in vivoand ensure the maximization of NH4

+ absorption by seedlings. Afterthat, seedlings were grown on 7.5 mM (NH4)2SO4 (HN) solutions.Samples (three biological replicates) of shoots and roots were takenseparately at 0, 4, and 12 h after the imposition of HN treatments,frozen immediately, and stored at −80 ◦C until RNA-Seq (Quantifi-cation) analysis.

2.2. Tissue ammonium (NH4+) content determination

Seedlings were harvested and desorbed for 5 min in 10 mMCaSO4 to remove extracellular NH4

+. Roots and shoots were sepa-rated and weighed, then transferred to 10-ml polypropylene tubesand frozen in liquid N2 for storage at −80 ◦C. Approximately 1 gof root or shoot tissue was homogenized under liquid N2 using amortar and pestle, followed by the addition of 6 ml of formic acid(10 mM) for the purpose of extracting NH4

+. 1 ml homogenate wascentrifuged at 12,000 rpm at 4 ◦C for 10 min. The supernatant wasthen transferred to 2-ml polypropylene tubes with 0.45-�m nylonfilters (Costar, Corning Inc., USA) and centrifuged at 12,000 rpmat 4 ◦C for 5 min. The resulting supernatant was analyzed by amodified o-phthalaldehyde (OPA) method (Balkos et al., 2010) todetermine total tissue NH4

+ content: 100 ml of OPA reagent wasprepared by combining 200 mM potassium phosphate buffer (com-posed of equimolar amounts of KH2PO4 and K2HPO4) and 3.75 mMOPA. Then, the solution pH was adjusted to 7 with 1 M NaOH. 2 mM2-mercaptoethanol was added into the solution one day before use.For sample testing, 200 �l aliquots of tissue extract were combinedwith 2 ml of OPA reagent. The color was allowed to develop in thedark for 15 min at room temperature, and sample absorbance wasmeasured at 410 nm.

2.3. RNA extraction, library preparation, and RNA sequencing ‘

Total RNA was isolated with TRIzol reagent (Invitrogen, USA)according to protocol of the manufacturer. Then, mRNA was puri-fied from total RNA using the Oligotex mRNA Midi Kit (Qiagen,Germany). After that, the total RNA samples were treated withDNase I to degrade any possible DNA contamination. Then, mRNAwas enriched using oligo (dT) magnetic beads. Mixed with thefragmentation buffer, mRNA was fragmented into short fragments.

Then, the first strands of cDNA were synthesized using randomhexamer-primer. Buffer, dNTPs, RNase H, and DNA polymerase Iwere added to synthesize the second strand. Double-strand cDNAwas purified with magnetic beads. End reparation and 3′-end

9 t Phy

sFf(Psm

2

adwerBggTl

F

uuTg

2

(ead

M

“s(Dfad

P

tb

2

pAndbs

6 L. Sun et al. / Journal of Plan

ingle nucleotide A (adenine) addition were then performed.inally, sequencing adaptors were ligated to the fragments. Theragments were enriched by PCR amplification. During the QCquality control) step, Agilent 2100 Bioanalyzer and ABI Step Onelus Real-Time PCR System were used to qualify and quantify theample library. Then, the library products were sequenced via Illu-ina HiSeqTM 2000.

.4. Mapping of RNA-Seq reads

Primary sequencing data, produced by Illumina HiSeqTM 2000nd called “raw reads”, were subjected to quality control (QC) toetermine if a re-sequencing step was needed. After QC, raw readsere filtered into clean reads, which would be aligned to the refer-

nce sequences. After filtering, the remaining reads, called “cleaneads”, were stored in FASTQ format (Cock et al., 2010). We usedowtie2 (Langmead et al., 2009) to map clean reads to referenceenes and used BWA (Li and Durbin., 2009) to reference the fullenome. Only uniquely mapped reads were used in the analysis.he gene expression (FPKM) levels were estimated using the fol-

owing formula:

PKM = 106C

NL/103

Given is the expression of gene A; C is the number of fragmentsniquely aligned to gene A; N is the total number of fragmentsniquely aligned to all genes; L is the number of bases in gene A.he calculated gene expression can be directly used for comparingene expression patterns among samples.

.5. Screening differentially expressed genes (DEGs) using NOISeq

NOISeq method can screen differentially expressed genesDEGs) between two groups. First, NOISeq uses a sample’s genexpression in each group to calculate log2 (fold change) M and thebsolute different value D of all pair conditions to build a noiseistribution model:

i = log2(xi1xi1

) and Di = |xi1 − xi2|

Second, for gene A, NOISeq computes its average expressionControl avg” in the control group and the average expres-ion “Treat avg” in the treatment group. Then, the fold changeMA = log2((Control avg)/(Treat avg))) and absolute different value

(DA = |Congrol avg- Treat avg|) are obtained. If MA and DA divergerom the noise distribution model markedly, gene A will be defineds a DEG. There is a probability value to assess how MA and DA bothiverge from the noise distribution model:

A = P(MA ≥ {M}&&DA ≥ {D})

Finally, we screened differentially expressed genes according tohe following default criteria: Fold change ≥2 and diverge proba-ility ≥0.8.

.6. Gene ontology (GO) function annotation

Gene Ontology (GO) classifications for a set of genes wereerformed by the Web Gene Ontology Annotation Plot (WEGO).ll genes in the rice genome were used as background for sig-

ificance testing. We first mapped all DEGs to GO terms in theatabase (http://www.geneontology.org/), calculating gene num-ers for every term. Then we used a hyper geometric test to findignificantly enriched GO terms in the input list of DEGs based on

siology 212 (2017) 94–104

‘GO:Term Finder’ (http://www.yeastgenome.org/help/analyze/go-term-finder), which is described as follows:

P = 1 −m−1∑

i=0

(Mi

)(N − Mn−i

)

(Nn

)

N is the number of all genes with GO annotation; n is the numberof DEGs in N; M is the number of all genes annotated to certainGO terms; m is the number of DEGs in M. The calculated p-valuegoes through Bonferroni correction (Abdi, 2007), taking correctedp-value ≤0.05 as a threshold. GO terms fulfilling this condition aredefined as significantly enriched GO terms in DEGs. This analysis isable to recognize the main biological functions that DEGs exercise.Blast2GO was further used to assign biological functions, cellularcomponents, and cellular processes to transcripts.

2.7. Kyoto encyclopedia of genes and genomes (KEGG) pathwayenrichment analysis

Mapped sequences were annotated against the Kyoto Ency-clopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/) to obtain enzyme commission (EC) numbers. Toobtain KEGG Pathway-Maps, the EC numbers were then mappedto the KEGG biochemical pathways. GO terms for genes from therice (version 6.1) reference genome were downloaded. Pathway-based analysis helps to further understand the biological functionsencoded by the various genes identified. KEGG (the major publicpathway-related database) was used to perform pathway enrich-ment analysis of DEGs. The formula used for calculations was thesame as that used for GO analysis:

P = 1 −m−1∑

i=0

(Mi

)(N − Mn−i

)

(Nn

)

Here, N is the number of all genes with KEGG annotation; n isthe number of DEGs in N; M is the number of all genes annotatedto specific pathways; m is the number of DEGs in M.

2.8. Quantitative real-Time PCR

Total RNA was extracted from shoots and roots harvested atthe specified time points with TRIzol reagent (Invitrogen, USA) andtreated with RNase-free DNaseI (Promega). Total RNA (2 �g) wasused for reverse transcription with M-MLV Reverse Transcriptase(Promega), and the cDNA samples were diluted two-fold. For qRT-PCR, triplicate quantitative assays were performed on each cDNAdilution with SYBR Premix Ex Taq (TaKaRa) and a CFX Managersequence detection system according to the following protocol:denaturation at 95 ◦C with 30 s for initiation, denaturation at 95 ◦Cfor 10 s, 40 cycles of amplification, annealing and extension at55 ◦C/60 ◦C for 30 s. Specificity of the amplification was checkedusing a melting curve performed from 65 to 95 ◦C, as well assequencing of the amplification. Three independent replicates wereperformed per experiment, and the means and corresponding stan-dard errors were calculated. Ubiquitin protein (UBI1) was used as anormalization control. Primer sequences were as listed in Supple-mentary Table S2.

2.9. Statistical analysis

All statistical analyses were conducted using SPSS version

13.0, and one-way ANOVA was performed with a homogeneity ofvariance test, followed by an LSD test to check for quantitative dif-ferences between treatments. P < 0.05 was set as the significancecut-off.

L. Sun et al. / Journal of Plant Physiology 212 (2017) 94–104 97

F d biomo f NH4

i ple siz

3

3

tarslb(

atgNraeiaT

3

aie8ewsphr

ig. 1. Rice phenotypic responses to different NH4+ concentrations. (a) The length an

f root and shoot under NH4+ stress [7.5 mM (NH4)2SO4]. (c) The expression levels o

ndicate standard errors of the mean of 30 plants. Mean values are shown for a sam

. Results

.1. Phenotypic response of rice to excess NH4+

In order to determine the highest NH4+ concentration rice could

olerate in our experimental conditions, we chose 1, 2.5, 5, 7.5, 10,nd 20 mM (NH4)2SO4, and monitored the changes in shoot andoot lengths and fresh biomass in 14-day-old plants. The resultshowed that root growth was retarded at 7.5 mM (NH4)2SO4. Shootength did not show any significant difference, but biomass inoth shoot and root were decreased by 50% at 7.5 mM (NH4)2SO4Fig. 1a). Thus, 7.5 mM (NH4)2SO4 was chosen as HN.

We measured changes in internal NH4+ concentration in shoots

nd roots prior to and following treatments at various differentime points, up to 48 h. Simultaneously, we analyzed the OsAMT1ene family by qRT-PCR, which is known to code for high-affinityH4

+ transporters (Sonoda et al., 2003). NH4+ concentrations in

oots and shoots began to increase between 4 h to 12 h in the root,fter which the NH4

+ concentration did not show further differ-nces (Fig. 1b). Meanwhile, transcripts of OsAMT1.1and OsAMT1.2n the root showed clear induction and reached the highest levelst 4 h, with OsAMT1.2 maintaining high levels up to 12 h (Fig. 1c).herefore, we chose 0, 4, and 12 h for RNA-Seq analysis.

.2. Identification of DEGs in response to NH4+ excess

We sequenced 18 samples using RNA-Seq technology, and theverage number of raw sequencing reads and clean reads are shownn Supplementary Table S1. The unique mapping ratio with refer-nce gene and the average genome mapping ratio were more than0% (Supplementary Table S1). To judge the significance of the genexpression differences and to define DEGs, two filtering criteriaere used in our data analysis: a two-fold or greater change in tran-

cript level between any two time points and a P-value <0.05. Fourrofiles (4S:4-h-Shoot/12S:12-h-Shoot and 4R:4-h-Root/12R:12--Root) were utilized to depict the transcripts of all DEGs in threeeplicates at 4 h and 12 h in both shoots and roots.

ass of shoot and root under different NH4+ conditions (HN). (b) NH4

+ concentrations+ transporter genes OsAMT1.1 and OsAMT1.2 in the root at 7.5 mM (NH4)2SO4. Barse of three replicates.

307 and 675 genes were up-regulated and 216 and 684 geneswere down-regulated, based on the analysis of 4S and 12S, respec-tively (Fig. 2a). In roots, 167 and 31 genes were up-regulated anddown-regulated, respectively, for 4R compared with 320 and 96genes for 12R (Fig. 2a). To identify both unique and common genesshowing differential expression patterns at the time points in bothshoot and root, numbers were calculated and presented using aVenn diagram (Fig. 2b). The results show that 367 and 153 DEGswere commonly induced on shoot and root, respectively. 5 DEGswere regulated in both shoot and root, demonstrating a progressivebiological process. Moreover, 133 and 932 DEGs were unique to theshoot, compared to 25 and 200 that were uniquely induced in theroot at 4 h and 12 h, respectively (Fig. 2b). In addition, GO enrich-ment analysis focusing on the categories of biological processeswas used to classify the biological functions of DEGs under HN. Asshown in Supplementary Fig. S1, the DEGs were classified into threemain categories including biological process, cellular component,and molecular function (corrected P-value <0.05). Cell, cell parts,and organelle under cellular component were the three most wellrepresented subcategories.

3.3. Roots and shoots exhibit distinct spatio-specific codes inresponse to excess NH4

+

We first used KEGG enrichment analysis to classify the cardi-nal biological processes and crucial regulatory genes in the rootunder the HN condition. As shown in Fig. 3a, the foremost pro-cesses that were engaged fell into the categories of plant-pathogeninteraction and hormone signaling pathways. The plant-pathogeninteraction category vis-a-vis the hypersensitive response (HR),especially programmed cell death (PCD) and cell wall enforcement,was up-regulated, followed by hormone signaling transduction,especially that involving the MAPK signaling pathway and defense-

related genes (Fig. 3a and Supplementary File S1). Concomitantly,transcripts involved in amino acid metabolism were up-regulated,mainly those which take part in lipid metabolism, includinga small subset of genes involved in fatty acid elongation and

98 L. Sun et al. / Journal of Plant Phy

Fig. 2. Numbers and Venn diagram display of DEGs in rice shoot and root under HNtreatments. (a) DEG number in shoot and root at 4 h and 12 h. (R value <0.05 andgenes with the regulation ratio log ≥ 2 or ≤−2 were selected). Blue: down-regulated.Red: up-regulated. (b) Venn diagram showing the overlapping number of DEGs inshoot and root at different time points. (Blue:4S, Pink:12S, Green:4R, Yellow:12R).4S:4-h-Shoot; 4R:4-h-Root; 12S:12-h-Shoot; 12R:4-h-Root.(For interpretation ofto

dIaoEaIbF

ognpmftcstcrt

genes were only induced at 4 h in the root, while 14 genes were

he references to colour in this figure legend, the reader is referred to the web versionf this article.)

esaturation in the plastid (Fig. 3a and Supplementary File S1).n addition, transcripts that map to glutathione, ascorbate, andldehyde metabolism were up-regulated as well, accompanied byxidative stress marker genes (Fig. 3a and Supplementary File S1).nergy metabolism pathways, such as oxidative phosphorylationnd fructose and mannose metabolism, were up-regulated notably.n addition, we found the transcripts of ABA and ET biosynthesis toe up-regulated in the root under HN (Fig. 3a and Supplementaryile S1).

In comparison with the root, the shoot must receive signalsn soil stresses from the root indirectly. As shown in Fig. 3b,enes related to chemical defense metabolism were the most sig-ificantly enhanced. Moreover, as in the root, genes related tohenylpropanoid biosynthesis were crucial within the secondaryetabolism category (Fig. 3b and Supplementary File S2). We

ound eight genes were constitutively up-regulated under HN. Allhese genes belong to peroxidases involved in flavonoid oxidation,innamyl alcohol dehydrogenase involved in lignin biosynthe-is, beta-glucosidase involved in a variety of processes such ashe release of phytohormones from inactive glycosides, and the

ytochrome P450 gene CYP98A/C3′H known to play an essentialole in the regulation of phenylpropanoid metabolism. Moreover,ranscripts related to key photosynthesis genes were increased

siology 212 (2017) 94–104

(Supplementary File S2). In accordance with the latter, genesin starch and sucrose metabolism and fructose and mannosemetabolism were up-regulated as well (Supplementary File S2).We also found increased expression levels of genes involved in theascorbate-glutathione antioxidant system and lipid metabolism,similar to what was revealed in the root. Interestingly, we foundtranscripts involved in GABA biosynthesis to be up-regulated 2.8-fold in the shoot under HN. Among hormone pathways, only ABAbiosynthesis genes were seen to be up-regulated in the shoot underHN (Supplementary File S2).

3.4. Temporal-specific codes in the shoot and root under excessNH4

+

With prolonged stress, root and shoot revealed temporal-specific codes. The top 20 KEGG pathways with the highestrepresentation of DEGs at the two time points in root and shootare shown in Fig. 4. In the root, the two most up-regulated DEGsbelonged to the plant-pathogen interaction pathway and to thedefense-associated hormone [gibberellin (GA), abscisic acid (ABA),ethylene (ET), jasmonic acid (JA), and salicylic acid (SA)] signalingcategory at 4 h of exposure, which was followed by increasing car-bohydrate and acid metabolism (Fig. 4 and Supplementary File S1).With increasing time of HN treatment, plant-pathogen defense sig-naling and ET, GA, and SA hormone signaling transcripts increaseda further 2-fold by 12 h, while ABA and JA did not change signifi-cantly. We also found antioxidant components, such as glutathionemetabolism and ascorbate and aldarate metabolism, to be up-regulated more than 2-fold by 12 h, accompanied by increasedenergy metabolism signaling related to galactose metabolism. Inaddition, transcripts relating to secondary metabolite biosynthesis,such as pyruvate, phenylalanine metabolism, and flavonoid biosyn-thesis, increased significantly (Fig. 4 and Supplementary File S1).

Maximum transcript numbers relating to the biosynthesis ofphenylpropanoids, such as lignin and flavonoids, were seen withincreasing time of NH4

+ stress at shoot (Fig. 4 and SupplementaryFile S2). Energy metabolism, most evidently starch and sucrosemetabolism, increased at 4 h under HN, and was especially vis-ible in amino sugar and nucleotide sugar metabolism. Otherthan this, energy metabolism, such as photosynthesis and gly-colysis/gluconeogenesis metabolism, was up-regulated by 12 h.Consistent with this, growth regulation genes were up-regulateddistinctly, while circadian rhythm gene expression was maximizedat 4 h (Fig. 4 and Supplementary File S2). Simultaneously, tran-scripts of ascorbate and lignin biosynthesis increased up to 7-fold.Additionally, glutathione metabolism increased (Fig. 4 and Sup-plementary File S2). We also found genes involved in carotenoidbiosynthesis enhanced distinctly at 4 h compared to 12 h, as shownin Fig. 4 and Supplementary File S2. Even though hormone signal-ing was not changed significantly, some transcription factors (TFs)were up-regulated (Supplementary File S2).

3.5. TFs (transcription Factors) in hormone regulation uponexposure to excess NH4

+

TFs are of special interest since they are capable of coordinatingthe expression of several or many downstream targets genes and,hence, entire metabolic and developmental pathways. As shownin Supplementary File S3, 32 TFs and 43 TFs were induced inthe root at 4 h and 12 h, respectively, and 17 TFs and 32 TFs inthe shoot at these time points. TFs also revealed spatio-temporalspecificity under HN. As shown in Supplementary File S3, three

expressed at 12 h. Likewise, five genes were only induced at 4 h inthe shoot, while 22 genes were expressed at 12 h. More importantly,most of the up-regulated TFs showed marked (>4-fold) changes in

L. Sun et al. / Journal of Plant Physiology 212 (2017) 94–104 99

Fig. 3. Spatial specificity in root and shoot under HN treatments. (a) The major defense and metabolism pathways in the root under HN. (b) The crucial metabolism pathwaysin the shoot under HN condition. R value <0.05 and genes with the regulation ratio log ≥ 2 or ≤−2 were selected.

100 L. Sun et al. / Journal of Plant Physiology 212 (2017) 94–104

F bolismo

tetMpdmpoetootgatg

3i

iuhrNTdLwia

ig. 4. Temporal specificity in root and shoot under HN treatments. The top 20 metar ≤-2 were selected. 4S:4-h-Shoot; 4R:4-h-Root; 12S:12-h-Shoot; 12R:4-h-Root.

ranscript abundance in response to HN condition. Some interestingxample are depicted in Table 1. The TFs uncovered mainly belongo the C2C2(Zn)CO-like zinc finger family and the AP2/EREBP and

YB families. Two of the G2-like (MYB-like) GARP, involved inhosphorous metabolism, abaxial cell identity, and photosyntheticevelopment, were induced under HN. Five MYB genes showedarked changes in transcript abundance under HN. AP2/EREBP

roteins,known to play important roles in plant growth and devel-pment throughout the plant life cycle especially in response tonvironmental stresses, were induced under HN (3-fold higherranscript level under HN). In addition, as shown in Table 1, manyf the TFs are involved in MAPK-based signal transduction. More-ver, 30 TFs related to protein degradation and modification viahe ubiquitin proteasome pathway were induced. To confirm theene expression profiling data obtained from RNA-seq, qRT-PCRnalysis of the selected TFs was performed as shown in Supplemen-ary Fig. S2. The differential expression observed with RNA-seq wasenerally confirmed with qRT-PCR data.

.6. Expression changes of homologous NH4+-related genes of rice

n response to excess NH4+

Many NH4+-toxicity-related genes in Arabidopsis have been

dentified, most of which are involved in root growth regulationnder excess NH4

+. To identify the expression changes of theiromologues in rice, we first identified the homologues genes inice using the Basic Local Alignment Search Tool (BLAST) from theational Center for Biotechnology Information (NCBI) as shown inable 2. We found not all homologous were induced under HN con-itions. Two of the Os VTC1 genes were up-regulated, while only

OC Os01g62840 was enhanced from 4 h to 12 h, as was the caseith Os AMOS1. The expression levels of two Os SLAH3 genes were

ncreased at 4 h and 12 h. However, Os CAP1 was only induced at 4 hnd decreased at 12 h, while Os AMT1.3 and one of Os Gln1.2 were

pathways were selected; R value <0.05 and genes with the regulation ratio log ≥ 2

only expressed at 12 h. Os AUX1 has five homologous genes in therice genome. Under HN, LOC Os03g14080 was induced at 4 h and12 h in the shoot, while LOC Os05g37470 was only expressed at12 h. Nevertheless, the expression levels of LOC Os10g05690 andLOC Os11g06820 increased at 4 h and 12 h with small changes.Unfortunately we did not find obvious change in LOC Os01g63770under HN. To confirm the gene expression profiling data obtainedby RNA-seq, qRT-PCR analysis was carried out as shown in Supple-mentary Fig. S3. The differential expression observed with RNA-seqwas generally confirmed with qRT-PCR data.

4. Discussion

The present study unravels the participation of the key signal-ing networks at the genome-wide level in rice in response to NH4

+

stress, which is hoped to promote general understanding of the reg-ulation of NH4

+ tolerance at the level of the whole plant. In total,1469 DEGs were up-regulated, while 1027 were down-regulated,under HN (Fig. 2b). GO and KEGG enrichment analysis revealedthat the up-regulated genes mainly fell into cellular function andmetabolic process categories, and gene transcription indicatedthat some important physiological or “housekeeping” functionsincreased significantly (Supplementary File S1 and S2).

4.1. Spatio-temporal specificity of gene expression in rice underNH4

+ toxicity

The root system is first in line when plants are challenged withstresses from the soil environment and, thus, must respond rapidlyto such stress signals to enable acclimation and tolerance. Among

the earliest defense responses in the root is the strengthening ofpre-existing physical and chemical barriers. Recent studies haveshown densely packed uniseriate sclerenchyma cells close to theroot apex strengthen their barriers to reduce the excessive NH4

+

L. Sun et al. / Journal of Plant Physiology 212 (2017) 94–104 101

Table 1Distribution and function of TFs under HN conditions. R value <0.05 and genes with the regulation ratio log ≥ 2 were selected.

Gene ID Annotation information log2Ratiounder HN

4RLOC Os09g26780/Os09g0439200 JAZ8/OsTIFY10c(jasmonate ZIM domain-containing protein 8)/C2C2(Zn)CO-like zinc finger

family/acts as a repressor of JA signaling in rice3 –

LOC Os03g27280/Os03g0390200 SnRK2/SAPK1(osmotic stress/abscisic acid-activated protein kinase1)/serine/threonine-protein kinase/MAPK signaling pathway

2.6 –

12RLOC Os07g03590/Os07g0126401 PR1(pathogenesis-related protein 1)/pathogenesis-related protein PRB1-3-like/MAPK

signaling pathway3 –

LOC Os06g13320/Os06g0241100 BRI1(protein brassinosteroid insensitive 1)/G-type lectin S-receptor-likeserine/threonine-protein kinase SD2-5/Signal transduction

6.8 –

LOC Os09g37330/Os09g0545300 SAUR36 (SAUR family protein)/auxin-responsive protein SAUR36/SMALL AUXIN UP RNA36/Signal transduction

5 –

LOC Os03g64260/Os03g0860100 1 ERF1(ethylene-responsive transcription factor 1)/AP2 domain containing protein/MAPKsignaling pathway

8.2 –

LOC Os05g33940/Os05g0410200 GID1(gibberellin receptor GID1)/Cell death associated protein/Signal transduction 3 –4SLOC Os03g58350/Os03g0797800 IAA14(Indoleacetic acid-induced protein 14)/Auxin-responsive protein

IAA14/auxin-activated signal transduction– 13.2

LOC Os03g08330/Os03g0181100 JAZ10 (jasmonateZIM domain-containing protein 10)/Protein TIFY 11b/jasmonic acidmediated signaling pathway

– 6

LOC Os02g27220/Os02g0471500 OsPP2C14(protein phosphatase 2C 14)/serine/threonine phosphatase activity/MAPKsignaling pathway

– 4.2

LOC Os02g10860/Os02g0203000 ABF(ABA responsive element binding factor)/gibberellic acid mediated signalingpathway/positive regulation of circadian rhythm

– 5.2

12SLOC Os05g43690/Os05g0512600 CRE1/AHK2/3/4(histidine kinase 2/3/4-cytokinin receptor)/hydrolase activity,polysaccharide

binding/Signal transduction– 10.6

LOC Os05g29030/Os05g0358500 OsPP2C48(protein phosphatase 2C 48)/serine/threonine phosphatase activity/MAPKsignaling pathway

– 5.6

LOC Os12g05660/Os12g0152800 TGA(transcription factor TGA)/protein tyrosine/serine/threonine phosphataseactivity/Signal transduction

– 5

LOC Os02g05630/Os02g0149800 OsPP2C10 protein phosphatase 2C 10)/serine/threonine phosphatase activity/PPM-typephosphatase/MAPK signaling pathway

– 4

LOC Os01g13740/Os01g0239000 ARR-B/OsGLK2(Probable transcription factor GLK2, Golden2-like protein 2)/MYB-type/Signaltransduction

– 5

LOC Os06g24070/Os06g0348800 ARR-B/OsGLK1(Probable transcription factor GLK1, Golden2-like protein 1)/MYB-type/Signaltransduction

– 4.4

LOC Os02g53200/Os02g0771700 CRE1/AHK2/3/4(histidine kinase 2/3/4-cytokinin receptor)/hydrolase activity,polysaccharidebinding/Putative beta-1,3-glucanase/Signal transduction

– 7.8

Table 2The expression levels of homologous genes in rice under HN treatments.

Gene name Gene ID Arabidopsis homologous gene 4S 4R 12S 12R

OsVTC1 LOC Os03g16150/Os03g0268400 VTC1(Glucose-1-phosphate adenylyltransferase) 2.6 −0.12 2.23 0.04LOC Os08g13930/Os08g0237200 0.6 0.45 1.72 0.29LOC Os01g62840/Os01g0847200 5.8 0.64 7.31 0.61

OsARG1 LOC Os01g32870/Os01g0512100 ARG1 (Chaperone DnaJ-domain protein) 0.34 0.02 0.46 −0.04LOC Os02g50760/Os02g0741100 1.2 0.02 0.78 −0.45

OsAMOS1 LOC Os03g57840/Os03g0792400 AMOS1 (Ethylene dependent gravitropism deficientand yellow green1/Peptidase M50 protein)

2.4 0.8 7.6 −2.61

OsAUX1 LOC Os01g63770/Os01g0856500 AUX1 (Auxin influx carrier protein 1) 0.86 0.11 0.44 0.12LOC Os03g14080/Os03g0244600 9.46 −1.02 9.47 −1.07LOC Os05g37470/Os05g0447200 −0.54 −0.99 2.63 −1.48LOC Os10g05690/Os10g0147400 7.14 0.002 6.72 −1.18LOC Os11g06820/Os11g0169200 3.78 −0.12 4.6 0.02

OsXBAT32 LOC Os02g54860/Os02g0791200 XBAT32 (Ring-type E3 ligase) 0.44 0.02 0.01 0.2OsETO1 LOC Os03g18360/Os03g0294700 ETO1 [Ethylene overproducer 1/tetratricopeptide

repeat (TPR)-containing protein]1.16 0.22 0.71 −0.03

OsDPMS1 LOC Os03g60939/Os03g0824400 DPMS1 (dolicholphosphate mannose synthase1/Nucleotide-diphospho-sugar transferasesprotein)

0.008 −0.41 0.27 0.02

OsCAP1 LOC Os03g55210/Os03g0759600 CAP1{[Ca(2 + )] cyt-associated protein kinase} 2.83 −0.99 1.4 0.51OsGln1.2 LOC Os02g50240/Os02g0735200 Gln1.2 (Cytosolic glutamine synthetase 1.2) 0.74 −0.26 2.96 0.48

LOC Os03g12290/Os03g0223400 0.07 −1.15 1.5 −0.54OsAMT1.3 LOC Os04g43070/Os04g0509600 AMT1.3 (Ammonium transporter 1.3) 1.19 0.29 5.6 1.2OsETR1 LOC Os03g49500/Os03g0701700 ETR1 (Ethylene response 1) 0.28 0.23 0.45 0.44OsSLAH3 Os05g0219900 SLAH3 (Slow anion channels/SLAC1-homolog

protein 3/S-type anion channel)8.32 1.12 3.68 0.86

LOC Os01g12680/Os01g0226600 2.65 0.54 6.48 −3.7OsTRH1 LOC Os07g48130/Os07g0679000 TRH1 (Potassium transporter protein) 1.15 0.48 −0.4 0.09

1 t Phy

ttfqerpbesspahpaucAAaartwiccaNcbpw

tttHNMpoItapssia(atpWAttdseaios

02 L. Sun et al. / Journal of Plan

ransport into the rice root (Ranathunge et al., 2016). Moreover,he ascorbate- glutathione cycle, one of the important processesor free radical detoxification under biotic or abiotic stress, is fre-uently engaged early (Domínguez-Valdivia et al., 2008; Wangt al., 2016). In our study, many genes involved in these defenseesponses were indeed up-regulated significantly (Fig. 3 and Sup-lementary File S1). These underscore that physical and chemicalarriers might play an important role for rice in the staging of thearly response to NH4

+ toxicity, as appears to be the case with othertresses. With increasing stress exposure time, many stress-relatedignaling pathways were activated, in particular those involvinghytohormones. It is known that ABA and ET signaling pathwaysre involved in Arabidopsis under NH4

+ toxicity in the root, but thisad not been examined in rice, as a model system of NH4

+-tolerantlants. We found ABA2, centrally involved in ABA biosynthesis,nd ACS and ACO, involved in ET biosynthesis, were up-regulatednder HN (Supplementary File S3). This suggests ABA and ET arerucial signaling molecules in rice, as they are in NH4

+-sensitiverabidopsis (Li et al., 2014). Correspondingly, we found the keyBA- and ET-response elements, ABFs and ERFs, were up-regulateds well (Supplementary File S3). Thus, the phytohormones ABAnd ET may be the major regulatory signaling molecules in theoot under HN. Furthermore, these signaling pathways belong tohe MAPK-signaling transduction pathway. MAPK signaling path-ays are present in both the cytoplasm and the nucleus and are

nvolved in a variety of crucial processes such as osmoregulation,ell growth, and differentiation (Mishra et al., 2006). Therefore, weonclude that the MAPK transduction pathway, regulated by ABAnd ET, plays a central role in the early stress response of rice toH4

+ stress in the root. The root not only reinforces its physical andhemical responses to retard damage incurred from NH4

+ stresselow-ground, but also activates ABA and ET signaling to the shoot,ossibly through the MAPK transduction pathway, and, in such aay, orchestrates defense signals sensitization and transmission.

An array of signals is typically generated when roots are exposedo an environmental stress. Some, but not all, are then conveyedo the shoot via the transpiration stream. Here, we found manyranscripts relating to MAPK signal transduction in the shoot underN (Supplementary File S3). This reveals that the transduction ofH4

+ toxicity signaling from root to shoot might occur through theAPK signaling pathway. Meanwhile, most transcripts involved in

henylpropanoid biosynthesis were up-regulated after 4 h and 12 hf NH4

+ exposure in the shoot under HN (Supplementary File S2).n higher plants, phenylpropanoid metabolism is the gateway forhe production of a great variety of secondary metabolites, suchs flavonoids, isoflavonoids, lignin, anthocyanin, phytoalexins, andhenolic esters, all of which are critical players in development,tructural protection, defense responses, and tolerance to abiotictimuli (Naoumkina et al., 2010). Induction of flavonoid synthesisn response to environmental stimuli can alter auxin transport tollow for adjustments in plant growth and development to stressBuer et al., 2007; Peer and Murphy, 2007). Nguyen et al. (2013)lso found ABA to modulate the biosynthesis of flavonoids and alterhe direction of polar auxin transport with the consequence of sup-ressing lateral root growth in plants suffering from drought stress.e also found the expression level of ABA2 transcripts, critical to

BA biosynthesis, up-regulated in the shoot under HN (Supplemen-ary File S3). Thus, we speculate that feedback signaling from shooto root in the early stages of NH4

+ exposure was via the MAPK trans-uction pathway, regulated by ABA. The shoot only perceives NH4

+

tress signals from the root indirectly, as it is not typically directlyxposed. Thus, intracellular physiological defense responses, such

s the biosynthesis of secondary metabolites, assume a primary rolen this process. These metabolites can regulate the polar transportf auxin and then regulate root growth in response to the imposedtress. Therefore, in the early NH4

+ stress defense process, the root

siology 212 (2017) 94–104

acts as a command center to activate and deliver stress-relateddefense signals, while the shoot engages in response and feedback.

4.2. Metabolism adjustments play an important role inNH4

+resistance

Environmental stress in plants also alters or disrupts metabolichomeostasis. In stressed conditions, plant metabolism must bemodified to produce compounds necessary to cope with and adaptto the stress (Shulaev et al., 2008). Hence, plants have to devotesignificant resources to stress defense (Heil and Baldwin, 2002).In agreement with this expectation, many transcripts in carbohy-drate metabolism pathways were up-regulated in both the root andshoot under HN (Supplementary File S1 and S2). However, rice,as an NH4

+-tolerant crop, must have especially well-respondingenergy metabolism with regard to this stress. Indeed, componentsof energy metabolism in response to HN increased considerably.Together with the amplification of the light reactions of photosyn-thesis, transcripts involved in the generation of ATP and NAD(P)H,which might be necessary to facilitate enhanced NH4

+ resistance,were up-regulated (Supplementary File S1 and S2). It is well knownthat adjustment of ATP formation and its utilization for synthesisof compatible solutes, and enhanced energy metabolism in general,are strategies plants use to cope with salt stress (Zhang et al., 2013;Zhao et al., 2013). In addition, elevated levels of amino acids andorganic acids in plant cells have been correlated with enhancedstress tolerance through the scavenging of free radicals and pro-tecting enzymes (Flowers and Colmer, 2015). Proline accumulationis a well-known measure to improve osmoprotection in particu-lar and is seen under a variety of stresses (Saxena et al., 2013).Congruously, transcripts in proline metabolism were increasedunder HN (Supplementary File S1 and S2). In addition, proline isalso involved in maintaining the NADP+/NADPH ratios requiredfor normal metabolism (Hare et al., 1998). However, in Arabidop-sis, proline accumulation has also been shown to have negativeimpacts, leading to cell toxicity under high temperature conditions(Lv et al., 2011). In rice, proline may serve as a positive factor inadapting to HN stress in the early stages of NH4

+ toxicity.

4.3. Role of phytohormones in NH4+tolerance

Phytohormones are molecules produced in very low concen-trations but work as chemical messengers to communicate cellularactivities and coordinate various signal transduction pathways dur-ing abiotic stress, integrating external as well as internal stimuli(Peleg and Blumwald, 2011; Vob et al., 2014). It has been reportedthat auxin, ET, and ABA are involved in NH4

+-responsive pathwaysin Arabidopsis (Li et al., 2014). Endogenous ABA levels increaserapidly, activate specific signaling pathways, and modify geneexpression to enable plants to survive under adverse environmen-tal conditions (Cutler et al., 2010). However, it has not been hithertoexamined whether ABA regulation is involved in response to NH4

+

toxicity in rice. In our study, the ABA biosynthesis genes ABA1, ABA2,and ABI1 were up-regulated in the rice root and shoot under HN(Supplementary File S3). Meanwhile, genes encoding Group A pro-tein phosphatases, PP2Cs, kinases, and SnRK2s were up-regulatedin the root and shoot under HN as well. In rice, differential expres-sion of several PP2C/SnRK2 was reported under abiotic stress andABA treatment (Xue et al., 2008), and they are known to serveas key regulators of ABA signaling in stresses such as desiccation(Komatsu et al., 2013). This shows that ABA plays a crucial rolein rice under the early stages of NH4

+ toxicity. In addition, EIN3

and ERF genes (AP2/EREBP), regulating stress response and devel-opment programs (Kazan, 2015), were up-regulated under HN,but only in the root (Supplementary File S3). ET often acts coop-eratively with JA and SA, and their biosynthesis, transport, and

t Phy

apssF(anlorbiTut

5

otNgamTyetarnahpNt

A

eCrg

A

i0

R

A

B

B

B

B

L. Sun et al. / Journal of Plan

ccumulation trigger a cascade of signaling pathways involved inlant defense (Matilla-Vazquez and Matilla, 2014). We found tran-cripts of MYC2, involved in JA synthesis, and PR1, involved in SAynthesis, were also up-regulated mainly in the root (Table 1).urthermore, Aux/IAA transcripts in the shoot were up-regulatedSupplementary File S1); auxin signaling is an integral part in plantdaptation to salinity stress (Fahad et al., 2015). Meanwhile, brassi-osteroids (BR) signaling is regulated by BRI1, and the expression

evel of BRI1 was increased at the two time points and increasedver time. However, BRI1 was only expressed in the shoot at 4 h,eturning to a normal level at 12 h (Supplementary File S3). It haseen shown that BRs are involved in plant growth and development

n Arabidopsis via the key regulator BRI1 (Jaillais and Vert, 2016).hese data uncovered the spatial specificity in hormone signalingnder HN. This provides important clues about the engagement ofhe key phytohormones in rice in response to NH4

+ stress.

. Conclusions

NH4+ profoundly affects various aspects of plant cell devel-

pment and metabolism. Plants differ greatly in their abilitieso process the NH4

+ ion and in their toxicity thresholds as soilH4

+ becomes elevated. Research into the responses to NH4+at the

enome-wide level in rice, a model system for NH4+ tolerance, will

llow for a greatly improved understanding of the physiological andolecular mechanisms underlying the NH4

+ toxicity syndrome.he present study provides a first spatio-temporal specificity anal-sis of the rice genome in response to NH4

+ toxicity. Our genexpression analysis results reveal key adjustments in the biosyn-hesis and regulation of phytohormones in rice in response to NH4

+

nd uncover the intricate metabolite regulation network at play inesponse to the ionic stress. Rice has the ability to redirect ions andutrients to prioritized processes and induce adequate metabolicdjustments for survival when challenged with high NH4

+. It is

oped that the molecular identification and characterization of thehenomenon of NH4

+ tolerance will help direct the improvement ofH4

+ tolerance in crop plants, many of which are highly susceptibleo NH4

+ toxicity at low to intermediate soil concentrations.

cknowledgements

This work was supported by grant of the National Natural Sci-nce Foundation of China (31430095, 31501821 and 31300210),hina Postdoctoral Science Foundation (2015T80593), and a Natu-al Sciences and Engineering Research Council of Canada (NSERC)rant to HJK (RGPIN-2014-05650).

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at http://dx.doi.org/10.1016/j.jplph.2017.02.06.

eferences

bdi, H., 2007. The bonferroni and Sidak corrections for multiple comparisons. In:Salkind, N.J. (Ed.), Encyclopedia of Measurement and Statistics. Sage, ThousandOaks, CA.

ai, L., Ma, X., Zhang, G., Song, S., Zhou, Y., Gao, L., Miao, Y., Song, C.P., 2014. Areceptor-like kinase mediates ammonium homeostasis and is important forthe polar growth of root hairs in Arabidopsis. Plant Cell. 26, 1497–1511.

alkos, K.D., Britto, D.T., Kronzucker, H.J., 2010. Optimization of ammoniumacquisition and metabolism by potassium in rice (Oryza sativa L. cv. IR-72).

Plant Cell Environ. 33, 23–34.

ittsánszky, A., Pilinszky, K., Gyulai, G., Komives, T., 2015. Overcoming ammoniumtoxicity. Plant Sci. 231, 184–190.

ritto, D.T., Kronzucker, H.J., 2002. NH4+ toxicity in higher plants: a critical review.

J. Plant Physiol. 159, 567–584.

siology 212 (2017) 94–104 103

Britto, D.T., Siddiqi, M.Y., Glass, A.D., Kronzucker, H.J., 2001. Futile transmembraneNH4(+) cycling: a cellular hypothesis to explain ammonium toxicity in plants.Proc. Natl. Acad. Sci. U. S. A. 98, 4255–4258.

Britto, D.T., Balkos, K.D., Becker, A., Coskun, D., Huynh, W.Q., Kronzucker, H.J., 2014.Potassium and nitrogen poising: physiological changes and biomass gains inrice and barley. Can. J. Plant Sci. 94, 1085–1089.

Buer, C.S., Muday, G.K., Djordjevic, M.A., 2007. Flavonoids are differentially takenup and transported long distances in Arabidopsis. Plant Physiol. 145, 478–490.

Chandran, A.K.N., Jung, K.H., 2014. Resources for systems biology in rice. J. PlantBiol. 57, 80–92.

Cock, P.J., Fields, C.J., Goto, N., Heuer, M.L., Rice, P.M., 2010. The Sanger FASTQ fileformat for sequences with quality scores, and the Solexa/Illumina FASTQVariants. Nucleic Acids Res. 38, 1767–1771.

Conrath, U., Beckers, G.J., Langenbach, C.J., Jaskiewicz, M.R., 2015. Priming forenhanced defense. Annu. Rev. Phytopathol. 53, 97–119.

Conrath, U., 2011. Molecular aspects of defense priming. Trends Plant Sci. 16,524–531.

Cutler, S.R., Rodriguez, P.L., Finkelstein, R.R., Abrams, S.R., 2010. Abscisic acid:emergence of a core signaling network. Annu. Rev. Plant Biol. 61, 651–679.

Domínguez-Valdivia, M.D., Aparicio-Tejo, P.M., Lamsfus, C., Cruz, C.,Martins-Louc ão, M.A., Moran, J.F., 2008. Nitrogen nutrition and antioxidantmetabolism in ammonium-tolerant and -sensitive plants. Plant Physiol. 132,359–369.

Fahad, S., Hussain, S., Bano, A., Saud, S., Hassan, S., Shan, D., Khan, F.A., Khan, F.,Chen, Y.T., Wu, C., Tabassum, M.A., Chun, M.X., Afzal, M., Jan, A., Jan, M.T.,Huang, J.L., 2015. Potential role of phytohormones and plantgrowth-promoting rhizobacteria in abiotic stresses: consequences forchanging environment. Environ. Sci. Pollut. Res. 22, 4907–4921.

Fan, X.R., Xie, D., Chen, J.G., Lu, H.Y., Xu, Y.L., Ma, C.H., Xu, G.H., 2014.Over-expression of OsPTR6 in rice increased plant growth at different nitrogensupplies but decreased nitrogen use efficiency at high ammonium supply.Plant Sci. 227, 1–11.

Flowers, T.J., Colmer, T.D., 2015. Plant salt tolerance: adaptations in halophytes.Ann. Bot. 115, 327–331.

Guan, M., de Bang, T.C., Pedersen, C., Schjoerring, J.K., 2016. Cytosolic glutaminesynthetase Gln1;2 is the main isozyme contributing to GS1 activity and can beup-regulated to relieve ammonium toxicity. Plant Physiol. 171, 1921–1933.

Hare, P.D., Cress, W.A., Staden, J.V., 1998. Dissecting the roles of osmolyteaccumulation during stress. Plant Cell Environ. 21, 535–553.

Heil, M., Baldwin, I.T., 2002. Fitness costs of induced resistance: emergingexperimental support for a slippery concept. Trends Plant Sci. 7, 61–67.

Jaillais, Y., Vert, G., 2016. Brassinosteroid signaling and BRI1 dynamics wentunderground. Curr. Opin. Plant Biol 33, 92–100.

Kazan, K., 2015. Diverse roles of jasmonates and ethylene in abiotic stresstolerance. Trends Plant Sci. 20, 219–229.

Komatsu, K., Suzuki, N., Kuwamura, M., Nishikawa, Y., Nakatani, M., Ohtawa, H.,Takezawa, D., Seki, M., Tanaka, M., Taji, T., Hayashi, T., Sakata, Y., 2013. Group APP2Cs evolved in land plants as key regulators of intrinsic desiccationtolerance. Nat. Commun. 4, 2219.

Kronzucker, H.J., Glass, A.D., Siddiqi, M.Y., 1999. Inhibition of nitrate uptake byammonium in barley. Analysis Of component fluxes. Plant Physiol. 120,283–292.

Kronzucker, H.J., Glass, A.D., Siddiqi, M.Y., Kirk, G.J.D., 2000. Comparative kineticanalysis of ammonium and nitrate acquisition by tropical lowland rice:implications for rice cultivation and yield potential. New Phytol. 145, 471–476.

Langmead, B., Trapnell, C., Pop, M., Salzberg, S.L., 2009. Ultrafast andmemory-efficient alignment of short DNA sequences to the human genome.Genome Biol. 10, 25–34.

Li, H., Durbin, R., 2009. Fast and accurate short read alignment withBurrows-Wheeler transform. Bioinformatics 25, 1754–1760.

Li, B.H., Li, G.J., Kronzucker, H.J., Baluska, F., Shi, W.M., 2014. Ammonium stress inArabidopsis: signaling, genetic loci, and physiological targets. Trends Plant Sci.19, 107–114.

Lv, W.T., Lin, B., Zhang, M., Hua, X.J., 2011. Proline accumulation is inhibitory toArabidopsis seedlings during heat stress. Plant Physiol. 156, 1921–1933.

Ma, X.L., Zhu, C.H., Yang, N., Gan, L.J., Xia, K., 2016. �-Aminobutyric acid additionalleviates ammonium toxicity by limiting ammonium accumulation in rice(Oryza sativa) seedlings. Plant Physiol. 158, 389–401.

Matilla-Vazquez, M.A., Matilla, A.J., 2014. Ethylene:Role in plants underenvironmental stress. In: Ahmad, P., Wani, M.R. (Eds.), PhysiologicalMechanisms and Adaptation Strategies in Plants Under ChangingEnvironment, 2. Springer Science+ Business Media, New York, pp. 63–94.

Mishra, N.S., Tuteja, R., Tuteja, N., 2006. Signaling through MAP kinase networks inplants. Arch. Biochem. Biophys. 452, 55–68.

Naoumkina, M.A., Zhao, Q., Gallego-Giraldo, L., Dai, X., Zhao, P.X., Dixon, R.A., 2010.Genome-wide analysis of phenylpropanoid defense pathways. Mol. PlantPathol. 11, 829–846.

Nguyen, H.N., Kim, J.H., Hyun, W.Y., Nguyen, N.T., Hong, S.W., Lee, H., 2013. TTG1-mediated flavonols biosynthesis alleviates root growth inhibition in responseto ABA. Plant Cell Rep. 32, 503–514.

Peer, W.A., Murphy, A.S., 2007. Flavonoids and auxin transport: modulators or

regulators? Trends Plant Sci. 12, 556–563.

Peleg, Z., Blumwald, E., 2011. Hormone balance and abiotic stress tolerance in cropplants. Curr. Opin. Plant Biol. 14, 290–295.

Ranathunge, K., El-Kereamy, A., Gidda, S., Bi, Y.M., Rothstein, S.J., 2014. AMT1;1transgenic rice plants with enhanced NH4

+ permeability show superior growth

1 t Phy

R

S

S

S

S

V

W

investigation of salt-responsive mechanisms in plant roots. J. Proteomics 82,230–253.

Zheng, X., He, K., Kleist, T., Chen, F., Luan, S., 2015. Anion channel SLAH3 functions

04 L. Sun et al. / Journal of Plan

and higher yield under optimal and suboptimal NH4+ conditions. J. Exp. Bot.

65, 965–979.anathunge, K., Schreiber, L., Bi, Y.M., Rothstein, S.J., 2016. Ammonium-induced

architectural and anatomical changes with altered suberin and lignin levelssignificantly change water and solute permeabilities of rice (Oryza sativa L.)roots. Planta 243, 231–249.

axena, S.C., Kaur, H., Verma, P., Petla, B.P., Andugula, V.R., Majee, M., 2013.Osmoprotectants: potential for crop improvement under adverse conditions.In: Tuteja, N., Singh, S.G. (Eds.), Plant Acclimation to Environmental Stress.Springer, NewYork, NY, pp. 197–232.

hulaev, V., Cortes, D., Miller, G., Mittler, R., 2008. Metabolomics for plant stressresponse. Plant Physiol. 132, 199–208.

onoda, Y., Ikeda, A., Saiki, S., von Wirén, N., Yamaya, T., Yamaguchi, J., 2003.Distinct expression and function of three ammonium transporter genes(OsAMT1;1-1;3) in rice. Plant Cell Physiol. 44, 726–734.

zczerba, M.W., Britto, D.T., Balkos, K.D., Kronzucker, H.J., 2008. Alleviation ofrapid, futile ammonium cycling at the plasma membrane by potassium revealsK+-sensitive and −insensitive components of NH4

+ transport. J. Exp. Bot. 59,303–313.

ob, U., Bishopp, A., Farcot, E., Bennett, M.J., 2014. Modelling hormonal responseand development. Trends Plant Sci. 19, 311–319.

ang, W., Li, R., Zhu, Q., Tang, X., Zhao, Q., 2016. Transcriptomic and physiologicalanalysis of common duckweed Lemna minor responses to NH4

+ toxicity. BMCPlant Biol. 16, 92.

siology 212 (2017) 94–104

Xie, Y.J., Mao, Y., Xu, S., Zhou, H., Duan, X.L., Cui, W.T., Zhang, J., Xu, G.H., 2015.Heme-heme oxygenase 1 system is involved in ammonium tolerance byregulating antioxidant defense in Oryza sativa. Plant Cell Environ. 38, 129–143.

Xue, T., Wang, D., Zhang, S., Ehlting, J., Ni, F., Jakab, S., Zheng, C., Zhong, Y., 2008.Genome-wide and expression analysis of protein phosphatase 2C in rice andArabidopsis. BMC Genomics 20, 550.

Yang, S.Y., Hao, D.L., Song, Z.Z., Yang, G.Z., Wang, L., Su, Y.H., 2015a. RNA-Seqanalysis of differentially expressed genes in rice under varied nitrogensupplies. Gene 555, 305–317.

Yang, H.Y., von der Fecht-Bartenbach, J., Friml, J., Lohmann, J.U., Neuhauser, B.,Ludewig, U., 2015b. Auxin-modulated root growth inhibition in Arabidopsisthaliana seedlings with ammonium as the sole nitrogen source. Funct. PlantBiol. 42, 239–251.

Zhang, Y., Fonslow, B.R., Shan, B., Baek, M.C., Yates III, J.R., 2013. Protein analysis byshotgun/bottom-up proteomics. Chem. Rev. 10, 234–294.

Zhao, Q., Zhang, H., Wang, T., Chen, S.X., Dai, S.J., 2013. Proteomics-based

in nitrate-dependent alleviation of ammonium toxicity in Arabidopsis. PlantCell Environ. 38, 474–486.